Small 2-way speakers with linear on-axis and power response characteristics (Scan Speak and SB Acoustics drivers). H&V off-axis measurements included

[kyuu]

Nevermind the schematic, wtf is this graph even?

???????????????????????????

 

[Roland68]

I'm not sure if this is the right place for this, but I'll give it a try. I asked Gemini AI to create something it's been dreaming of since the Mechano23 kits were created: a bass module, without modifying the Mechano23 crossover, just like CSS did in its iconic bass module (https://www.css-audio.com/online-store/3-Way-Bass-Module-Add-On-Pair-p350688026).
What the AI did was, of course, theoretical, an attempt to combine components based solely on their physical properties. Now it's time for the voice of the practitioners—your voice, dear forum members—before I take the plunge and buy the parts the AI listed. Here are some inputs. I wanted the cost of the components to be similar to the Mechano23. The dimensions were also supposed to be similar.
I want the bass to extend lower and shake the windows more. Do I want that much?

Box, 18mm plywood:
Side 300 mm×650 mm
Up/Down 190 mm×300 mm
Front/Back 154 mm×650 mm

Speaker: Peerless SLS-8 (P830667)
Passive radiator: SB20PFCR-00 with mass 16g (F3≈31 Hz)
Crossover, second order 150Hz 12dB (C=270uF, L=8.2mH)

View attachment 497756

Finally, I asked for a frequency graph...
View attachment 497758

Do you think "it will work"?

Everything about this is as bad and wrong as it could possibly be.
Anyone who relies on AI's predictions in the audio field is doomed. This is proven time and again.

The crossover for the subwoofer is unusable.
The inductor for the low-pass filter must be in the signal path to the woofer, not in parallel.
The capacitor for the low-pass filter must be in parallel with the woofer, not in series.
A coil specification for such a coil without a resistance value is useless.
An inductor in the signal path without considering the saturation current is unusable.
The passive radiator is incorrectly dimensioned.
The subwoofer driver's specifications indicate a significant impedance mismatch with the Mechano23 when using a passive crossover (too quiet).

The lack of a high-pass filter for the Mechano23 reduces its sound quality to that of a cheap party speaker. Furthermore, it could be damaged at high volumes because the low frequencies are not filtered.

A subwoofer enclosure made of 18mm plywood? You can do it, but you can also just forget about it.

High-pass and low-pass filters are necessary for a halfway decent result.
With the expected cutoff frequency of 60-150 Hz, high-quality components are already so expensive and inflexible that an active crossover with an additional amplifier is definitely the better choice. A stereo amplifier with 2 x 100-150 watts and an active crossover with 4 output channels can be found for as little as €200 and makes much more sense.

 

[Bwaslo]

Nevermind the schematic, wtf is this graph even?

???????????????????????????

It's a plot of FPecnsse vs. Frequency. Problem with that??

 

[Roland68]

This one is listed as 7 x 18mm but I'm skeptical if that is true :

Jantzen Audio 27uF 400V Cross Cap Crossover Capacitor

Jantzen Audio 27uF 400V Cross Cap Crossover CapacitorProduct FeaturesDimensions: (Ø 7mm - L: 18mm)The Cross Cap is offering a high-level of performance and quality at a very reasonable price point. A great multipurpose entry-level audiophile capacitor. A single foil (standard type MKP/PP foil)...

www.parts-express.com

and its more expensive @~ $16 each.

The correct dimensions are probably 32 x 58 mm.

You could take a look at industrial-grade MKP capacitors from Kemet and Epcos/TDK. 27uF 800V capacitors start at around €5, and 1000-1100V capacitors start at around €10.
Significantly better electrical properties and cheaper. I've been using them for over 20 years.

 

[sebastian k]

I re-measured the enclosure and the volume is very close to 4l with very generous internal bracing (I used 12mm birch plywood for making the speaker lighter) and crossover included. The overall gross volume of the box is 4,9l. The filling is 25mm thick polyester wool on every enclosure wall in the woofer compartment and the tweeter compartment stuffed full. I placed the crossover behind the woofer, but it was a snug fit (about 1,5 cm room between the magnet and crossover).

I didn´t want to throw my old Wharfedales away, so I made them new enclosures with the same style too and recapped the crossovers. The sound is nowhere near as good, but I guess they work ok in atmos height channel.
View attachment 491841

Thanks a lot for your clarifications. They helped a lot. And sorry, I did not reply sooner - I was quite busy in December...

Internal stuffing seems to make the difference here whereas box volume doesn't seem to make too big of a difference. Another thing that might have helped is that I improved my temporary enclosure. I sealed up some leaks and added regidity - basically it's the front baffle and a bunch of wood clamps at this time ;-)

Anyways - results are not perfect, but probably close enough and definitely better than my preveous attempt. Red is the original Mechano 23, blue is the sealed version.

I do have the original baffle wall and an internal depth of 11.5 cm with 19mm MDF. So the volume is just under 4 l. The effecive internal volume is smaller of course because of the drivers and the crossover.

Stuffing is one layer of this visaton polyester wool on the side walls and the top and bottom walls.
Additional stuffing behind the tweeter as you proposed did make things worse.

 

[glampersand]

Hey Y'all, it's been something like 20 years since I last built a set of speakers and I'm thinking of doing a set of these to be my new main set. (My last build was a set of Selah Audio RCR3, which I know the sealed version didn't measure well here but I've loved my ported set — I'd still be using them they fit in my space better!) Planning them to take the place of an old set of AV123 ELT525 in a roughly 25' x 14' living room, paired with a Wiim amp and SVS PB2000 sub. I know they're a bit small for that kinda space but we don't listen at particularly high levels and the AV123 set are also just 5" woofer bookshelfs that have been fine output-wise when combined with the sub crossed round 80hz.

A quick question before I put in the orders. Right now I've got the crossover parts from the OP in a couple different shopping carts, and I'd really like to try a rear mounted passive radiator, mainly because I've not built with one before and it seems fun to try. Through this 40 page thread and the review thread there's a bit of talk about alternate crossover options — is there anything I should definitely look at doing different than in the OP? I've got all the right values tracked down so that's not a concern, but I don't know enough about crossover design to have a sense of what's smarter folks just noodling on on things that might add a hair of improvement vs obvious upgrades.

 

[Ktacos]

Hey Y'all, it's been something like 20 years since I last built a set of speakers and I'm thinking of doing a set of these to be my new main set. (My last build was a set of Selah Audio RCR3, which I know the sealed version didn't measure well here but I've loved my ported set — I'd still be using them they fit in my space better!) Planning them to take the place of an old set of AV123 ELT525 in a roughly 25' x 14' living room, paired with a Wiim amp and SVS PB2000 sub. I know they're a bit small for that kinda space but we don't listen at particularly high levels and the AV123 set are also just 5" woofer bookshelfs that have been fine output-wise when combined with the sub crossed round 80hz.

A quick question before I put in the orders. Right now I've got the crossover parts from the OP in a couple different shopping carts, and I'd really like to try a rear mounted passive radiator, mainly because I've not built with one before and it seems fun to try. Through this 40 page thread and the review thread there's a bit of talk about alternate crossover options — is there anything I should definitely look at doing different than in the OP? I've got all the right values tracked down so that's not a concern, but I don't know enough about crossover design to have a sense of what's smarter folks just noodling on on things that might add a hair of improvement vs obvious upgrades.

You don't need a passive radiator if you plan on using them with subs. The PR's that would work aren't exactly cheap either.

 

[glampersand]

Planning to keep them for a long time and in 5-10 years when the kids are a bit less of a menace I’ll probably rearrange and use these on their own somewhere. I’d been looking at the SB 5x8 which is more expensive than a tube but if it’s another 20 years before I build another set may as well live a little.

 

[ChillHippo]

Long time lurker, first time poster. I finished a test unit the other day and finally got to measuring, design mirrors the reference with one primary change to the enclosure shape and port location. It is mounted to the wall using a french cleat with a rear 1" cavity to give room for the cleat and in-wall cable, making it a fantastic Theater side and rear designs.

Used the port and PCB from wineds (thank you!). 1/2" mdf, doubled for the front, face of it has the same dimensions as the reference post then widens out to net ~8L, port on the bottom roughly center.

Sharing as someone might be interested in the general design of it and results.

Seating position in non-treated room, the drop at 95, 285, and 365hz seem room related as they change with different speaker locations.

Off to build the next 3 and make this one look pretty.

 

[codercodi]

Most of my parts we're surprisingly huge. I had to rearrange the parts a little but too. But make sure to keep the Inductor in the correct position/ alignment as in the original design. Else they will effect each other.

Do you have pics of your arrangement? Would love to take a look at them.

 

[klapper]

Do you have pics of your arrangement? Would love to take a look at them.

Bit late but that's the only picture i have. I used the 3D printed plate for the crossover. But it is quite wobbly and soft.

Edit: this is zoomed because the images seems to get compressed:

 

[Roland68]

Bit late but that's the only picture i have. I used the 3D printed plate for the crossover. But it is quite wobbly and soft.View attachment 503408View attachment 503409
Edit: this is zoomed because the images seems to get compressed:View attachment 503410

The fact that the plate is soft is partly due to its structure.

For maximum hardness, PC or PC GF would be suitable materials.
Closed surfaces on both sides with 15-25% infill, adding only the necessary holes. Alternatively, a grid of holes with 7-10mm spacing can be used.
For even greater stability, the edge can be reinforced with a profile, and a 3D structure can be used on the top and/or bottom, similar to what is used in aluminum or magnesium die casting. This can also be designed as a part divider or snap-in fastener.

For anyone planning to print something like this: CF filaments are unsuitable due to their potential conductivity, unless adequate insulation is used.